d as representative of achievement. The Stanford Achievement Test was the first of the large-scale publishers' tests, first published in 1923 and followed by the American College Testing Program (ACT), the Scholastic Aptitude Test (SAT) and others. By the 1950s, the Bloom Taxonomy emerged, in which teaching by objectives (many related to test performance) was justified and test quality was raised to reflect more than simple rote-memory learning. These tests were generally norm-referenced - a concept to be examined below.
In the 1930s, the work of Ralph Tyler introduced the notion of teaching and testing by instructional objectives. As Haladyna, et al (1998), point out, in the 1970s this method was linked to criterion-referenced rather than norm-referenced tests such as the ACT and SAT. In the latter half of the twentieth century, standardized testing was affected by three factors: 1) changing demographics due to immigration; 2) technological challenges emerging from the Cold War and the advent of the computer age; and 3) racial inequality, which has fostered charges that most standardized tests are inherently biased against minority groups (Haladyna, et al, 1998).
Lewis (1999) reported that the publication of A Nation at Risk and the Goals 2000 initiative refocused attention on issues related to accountability and standards. States set new standards and assessments to be applied to all students. Concerns regarding the viability of all types of standardized tests have continued to surface, however, as demands for national standards and national tests are also being expressed (Lewis, 1997).
What tests actually measure or attempt to measure, according to Sireci, DeLeon, and Washington (2002), is "knowledge." Standardized tests of all types are an attempt to level the playing field for all students by ensuring uniform test content and administration conditions as well as scoring strategies. Nevertheless, Sireci, et al (...